Understanding the Message Passing in Graph Neural Networks via Power Iteration Clustering

Abstract

The mechanism of message passing in graph neural networks (GNNs) is still mysterious. Apart from convolutional neural networks, no theoretical origin for GNNs has been proposed. To our surprise, message passing can be best understood in terms of power iteration. By fully or partly removing activation functions and layer weights of GNNs, we propose subspace power iteration clustering (SPIC) models that iteratively learn with only one aggregator. Experiments show that our models extend GNNs and enhance their capability to process random featured networks. Moreover, we demonstrate the redundancy of some state-of-the-art GNNs in design and define a lower limit for model evaluation by a random aggregator of message passing. Our findings push the boundaries of the theoretical understanding of neural networks.

References in corpus (9)

• Semi-Supervised Classification with Graph Convolutional Networks
• Simplifying Graph Convolutional Networks
• GNNExplainer: Generating Explanations for Graph Neural Networks
• Predicting multicellular function through multi-layer tissue networks
• Predict then Propagate: Graph Neural Networks meet Personalized PageRank
• Scaling Graph Neural Networks with Approximate PageRank
• Understanding the Representation Power of Graph Neural Networks in Learning Graph Topology
• GraphLIME: Local Interpretable Model Explanations for Graph Neural Networks
• Interpreting and Understanding Graph Convolutional Neural Network using Gradient-based Attribution Method

Cited by in corpus (3)

• Generating Post-hoc Explanations for Skip-gram-based Node Embeddings by Identifying Important Nodes with Bridgeness
• Towards explainable message passing networks for predicting carbon dioxide adsorption in metal-organic frameworks
• Does your graph need a confidence boost? Convergent boosted smoothing on graphs with tabular node features